The field of microfluidics can be broadly defined as related to the manipulation of small fluidic volumes, typically those that are less than one milliliter. The concept that using microfluidic volumes in clinical assays would represent a significant improvement over conventional methods dates back to the late 1960's and early 1970's. While those working in the field appreciated that small volume analyses would lead to portable instruments that required less laboratory space, improved precision and accuracy, increased throughput, reduced cost, and compatibility with true automation, the systems developed at that time were simple and did not realize the desired advantages of microfluidics. One early system was based on the analysis of 1-10 μl of sample and 70-110 μl of reagents that were transferred and mixed in a small centrifugal rotor (Anderson, N. (1969). Computer interfaced fast analyzers. Science 166: 317-24; and Burtis, C et al. (1972). Development of a miniature fast analyzer. Clinical Chemistry 18: 753-61). That rotor is likely the first reported microfluidic biochip.
In 1990, the theoretical foundation of the field of microfluidics was further characterized by Manz, who coined the term “miniaturized total chemical analysis system” or “μ-TAS”, to define what he considered should be the next generation of microfluidic devices. Manz proposed that such devices should be capable of performing all required sample handling steps. Manz' stated objectives helped establish the major goal in the modern era of microfluidics: to deliver a total, that is to say, fully integrated, analysis systems capable of performing a complex series of process steps from the insertion of a sample to the generation of a result without operator intervention. It is the promise of integrating a complex series of sample manipulations and process steps that has led to the widespread adoption of the phrase “laboratory on a chip.”
Work on centrifugal systems continued after the first early efforts and resulted in a commercial centrifugal system for sandwich immunoassays that used a microfluidic compact disc drive. The system was not fully integrated, however; and more accurately should be termed a workstation in which reagents were transferred to and from the compact disc robotically. Furthermore, the immunoassays were quite simple, based on passing a sample through a capture column. Although the system used microfluidic volumes to reduce reagent costs, it delivered neither on the promise of microfluidics to provide a fully-integrated system nor on the promise of microfluidics to integrate a complex series of sample manipulations and process steps.
Similarly, other workers fabricated a spinning biochip in poly(methyl methacrylate) [PMMA] by soft embossing to fractionate plasma from whole blood. There have been applications of spinning microfluidic biochips, for measurement of various analytes of interest in a blood sample. However, the devices have been developed for simple analytic tests, for example establishing the amounts of analytes of interest in a blood sample. The samples are not subjected to a complex series of sample manipulations and process steps, and even after almost two decades of development, can still have significant reliability problems correlating with established conventional assays.
In general, CD-based biochips have major drawbacks and limitations. First, they do not enable sufficient process complexity. These biochips are limited to relatively simple processes such as certain cell separations (serving as replacements for conventional centrifugation instrumentation) and for high throughput immunoassays (serving as replacements for the mixing and incubation of conventional assays). Second, the use of centrifugal force as the driver of fluidic transport is limiting. For example, the requirement to rotate the biochip places profound limitations on the approach to sample handling—either the sample must be introduced indirectly (requiring either manual intervention or additional instrumentation) or directly (e.g. a blood collection tube or swab would need to be subjected to centrifugation as well). Furthermore, process flow in spinning biochips proceeds unidirectionally, and the radius of the biochip limits the area available for sample process steps to take place (one of the factors that limits process complexity).
Concomitantly, alternatives to CD-based biochips have been studied. A number of groups have worked on microfluidic biochips based on microtiter plates using SBS (Society for Biomolecular Screening, Microplate Standards Development Committee, ANSI/SBS1, Danbury, Conn., 2004) standards. The SBS developed a series of published standards for microtiter plates that include footprint (127.76 by 85.48 mm, 10,920.9 mm2), height, and bottom outside flange dimensions and well positions. These standards have been incorporated by commercial manufacturers and academic groups, regardless of the fabrication process utilized. Microfluidic microtiter plates have been developed for a number of processes, including DNA purification and protein crystallization. These biochips require sample pre-processing prior to introduction, perform only two steps of the DNA purification process, and do not perform any analysis of the DNA product of the process.
Similarly, other groups have used rectangular microfluidic plates based on SBS standards for tasks such as the mixing of nanoliter volume reagents, medium exchange during live cell microscopy, and dispensing cells and reagents. Still other workers have adapted rectangular biochips to be used in robotic systems for the high-throughput dispensing of reagents into wells. Another group developed a rectangular biochip for protein detection by biological signal amplification in a well-in-a-well device.
In the progression of biochip development, a number of “integrated fluidic circuit” products based on the use of microfluidic features in an SBS format have been commercialized. For example, a biochip performs single nucleotide polymorphism (SNP) genotyping using an array that can interrogate 48 samples for each of 48 SNPs. The process is conducted by first preparing purified DNA and reaction mixes outside the microfluidic plate, placing the plate on a controller, priming the plate, loading and mixing reagents over 45 minutes on the plate, placing the plate on a thermal cycler for PCR, and transferring the plate to a fluorescence detection instrument. These biochips require sample pre-processing prior to introduction, require several manual steps during on-chip sample processing, do not incorporate reagents, and do not perform any analysis of the DNA product of the process.
A similar commercial approach measures platelet adhesion using a well plate microfluidic technology that integrates micron scale flow cell devices into SBS-standard well plates. The process requires a number of pre-processing steps, including manually introducing a protein of interest to coat the microfluidic channels, perfusing the channels, washing, manually preparing the cell sample of interest, adding the cell sample to the microfluidic channels, and placing the plate onto a workstation for further processing and analysis. A similar approach has been applied to studying wound healing by exposing epithelial cells to a variety of compounds. These biochips require extensive sample pre-processing prior to introduction.
Much of the microfluidic biochip prior art is based on the use of materials such as silicon or glass to fabricate biochips. However, silicon and glass biochips are prohibitively expensive to fabricate for high volume commercial applications (costing up to thousands of dollars for a single biochip) and are therefore impractical to be single-use disposables. Moreover, the need to reuse these biochips leads to problems with run-to-run contamination (an issue for human identity and clinical diagnostics), problems with instrumentation complexity if the biochips are prepared for reuse in the instrument, and problems with logistics if the biochips are prepared for reuse outside the instrument.
In order to realize the unfulfilled potential of microfluidics, microfluidic biochips and systems that are capable of performing a complex series of processing steps for one or more samples in parallel in the setting of a fully-integrated, sample-in to results out system in which there is no requirement for operator manipulation is needed. However, the biochips developed to date perform only a subset of the steps required for analysis, cannot perform complex series of processing steps, and are not capable of producing analytical results on a single device. Furthermore, these biochips and systems use materials and methods ill-suited to mass production.